Molecular engineering of antibodies specific for tumor-associated antigens will be undertaken with the goal of producing agents optimized for clinical radioimmunotherapy and radioimmunoimaging. The carcinoembryonic antigen (CEA) system remains the primary focus of the work. Humanized versions of the anti-CEA T84.66 antibody have been developed and evaluated, and further engineered fragments will incorporate humanized anti-CEA variable regions. Proposed work focuses in three areas: format, new target antigens, and immunofusions. (1) Recombinant antibody fragment formats will be further investigated. An intermediate MW fragment, the minibody (scFv-CH3 dimer, 80 kDa), which previously demonstrated rapid, high-level targeting of tumors in preclinical studies, is being evaluated as a clinical imaging agent. Preliminary animal therapy studies of 90Y minibody suggest this format may deliver sufficient dose to impact tumor growth. One additional series of fragments will be generated, based on the slightly larger scFv-Fc format (110 kDa), which displays slower kinetics and clearance similar to intact antibody. A series of mutations in the FcRn binding site on CH2 and CH3 will be evaluated for their impact on blood clearance and tumor uptake of the recombinant fragments, in order to obtain candidate fragments with performance intermediate between intact and minibody. Performance will be compared with existing fragments for selection of appropriate candidates for clinical radioimmunotherapy studies. (2) Knowledge gained in the CEA system will be extended to two additional targets in solid tumors: Her2 and prostate stem cell antigen (PSCA). Diabodies (scFv dimers, 55 kDa), minibodies, and scFv-Fc fragments will be produced from anti-Her2 and anti-PSCA antibodies provided by Dr. Michael Press (USC) and Dr. Robert Reiter (UCLA). Tumor targeting of radiolabeled fragments will be evaluated in athymic mice bearing appropriate antigen-positive and control tumors. The goal is two-fold: extension of our knowledge relating antibody format to in vivo performance, and provision of new candidate agents for clinical radioimmunoimaging and radioimmunotherapy studies. (3) Immunofusions will be explored for their ability to enhance radioimmunoimaging and radioimmunotherapy. An anti-CEA-IL-2 immunocytokine will be produced to locally increase vascular permeability, and an anti-CEA-gamma-interferon fusion will be generated to locally increase CEA expression of tumor cells. Preadministration of these agents should increase tumor accretion of subsequently administered radiolabeled antibodies. Protein engineering provides a powerful approach for tailoring antibodies for use as radiopharmaceuticals in patients.